Pandora’s Facebook Box

(image) I flew to Detroit today, and thankfully Delta had wifi. Since I'll be speaking at a GM conference later in the week, and the fine folks from Pandora will be there, among others, I went and checked in on the site, which I'll admit I haven't visited in…


(image) I flew to Detroit today, and thankfully Delta had wifi. Since I’ll be speaking at a GM conference later in the week, and the fine folks from Pandora will be there, among others, I went and checked in on the site, which I’ll admit I haven’t visited in some time (I still consume music the old fashioned way – I buy CDs and rip them to iTunes). Now, the theme of GM’s internal conference is all about “the app economy” and fortunately, lately I’ve found myself thinking a lot about this samesaid phenomenon. Given that, allow me to digress. As usual, I have no idea where this is going, but at least I know where it’s going to start: With my first visit to Pandora in some time.

Here’s what happened. Pandora has done a “deep integration” with Facebook since my last visit (yeah it’s been a while), meaning that when I showed up (and was logged into Facebook already), Pandora went ahead and filled out my profile using Facebook data. To the site’s credit (and I hope based on some terms of service from Facebook), the service notified me of this, and asked me if using my Facebook profile was OK.

Now, you may recall the kitty-with-a-ball-of-yarn that is my Facebook account. In short, it’s a tangled mess, and I’m at a loss around what to do about it. Short version: I said yes to the first 5000 folks who asked to be my “friend” and found myself with a pretty useless “social graph.” I’ve tried a few times to remedy the situation, but Facebook ain’t making it easy. The service wants you to be who you already are, not who you might want to become, that much is obvious. And who I already am on Facebook is a not-so-hot mess.

So…now I’m faced with importing this samesaid mess into Pandora, a place I was hoping to craft in the image of my own musical tastes. Do I click “OK”, or do I do the sensible thing, ditch the Facebook integration, and start from scratch? I mean, I have no idea how Pandora was *actually* going to use the data it got from Facebook, did I? Obviously the sensible thing was to be cautious, and click No F’in Way.

Of course I clicked Go Ahead, Use the Mess. Because, in the end, all I wanted to do was get to the music, consequences be dammed. Sure, I had no idea how or what Pandora was really going to do with my Facebook data, but honestly, I kind of didn’t care. I figured if it sucked, I’d find a way out. Right? (Actually, yes, you can undo the connection in settings.)

But connecting to Facebook got me thinking. First off, I wondered if Pandora even knew what do to with my “social graph” – given it has no rhyme or reason, and with 5000 or so connections, should Pandora really want to Go Deep, it’d probably melt a few CPUs down at the Music Genome project. And second, it made me wonder whether, had I chosen instead to do the work at Pandora, building my own profile from scratch…well had I done that, I’ll tell you this: I’d sure as hell like to import THAT profile into Facebook, and make THAT profile who I am up in ZuckerLand. Because it sure would reflect my identity a heckuva lot better than Facebook does at the current moment.

Hmmm. Now there’s an idea. What if I could take all that declaration of who I am that I do out on the “rest of the web”, and somehow drive that back INTO Facebook, in such a way as to shift Facebook’s understanding of who I am in a way that I controlled? And what if I could do that over and over, creating all sorts of different identities, ones I could mix and match on a whim, or a mood, or a social instance? Wouldn’t that be cool? I mean, if I could start all over, from scratch, I think I’d like to start at a place like Pandora, build a profile of who I am, and then import that profile (sort of like a piece of digital clothing) into a place like Facebook. Starting at Facebook, in a way, seems backassward. I’m not who I say I am, or who I say my friends are, one time on one platform built just for declaring my identity.

I’m what I do, in context, and that context shifts based on any number of axes – who I’m sharing with, social frame (professional? personal? familial? commercial? intimate? public? etc.), hell, it even shifts with my mood. And it sure as heck shifts over time. (I think this is what Eric was referring to when he joked that we should all have the right to get a new identity after college).

Increasingly, I’m frustrated with a world that wants me to be one thing – one profile, one easily structured dataset, one ring to rule them all. This just ain’t the way the real world works. It’s what I was getting at when I penned “Identity and the Independent Web” last year, and it’s a piece of yarn I’ll continue to pull at, mess be dammed. I want to be able to push data back into Facebook, such that Facebook changes who it thinks I am, and I want to be in control of that process.

In other words, I’d love to be able to tell Facebook, I’m feeling Pandorish right about now…show me what you got for me now?

And I predict that day will come. If not with Facebook, then with a platform that understands me better, one I’ll be more than happy to inhabit.

Am I crazy, or just too early? Tell me what you think.

19 thoughts on “Pandora’s Facebook Box”

  1. Microsoft recognized years ago (if it’s live today it was spec’ed years ago) that a human’s identity on the web is an amalgam of their identity around the web.

    Many of the capabilities you mention are in action at I don’t expect Microsoft to put the pieces together fast enough, but they do have the pieces.

  2. Well, I did not use Facebook connect to sign up for this comment

    I think identity management is FB real business. I am ok with FB playing with my identity within FB, a bit like I o not mind friends teasing me at a pub. But it is entirely different out of the pub.

    The point is convenience. It’s crazy that webshould abandon mgmt of our identity (I lové the I am what I do in context) for convenience’s sake.

    The tide Will turn probably soon

  3. Of course you are crazy.


    This is the fundamental problem of Facebook.

    It is a great repository of all of your digital information, but because it encompasses all of the different pieces of your identity perhaps it does a great job for you as a whole – but not the you of the moment.

    So if you are a little like Sybil and have multiple overlapping personas depending on the situation (and who doesn’t) how do you tease the proper one out in Facebook? you can’t.

    And unfortunately what everyone on Facebook gets is all the personalities – not the ones we necessarily want to show.

    So your Pandora personality might not mesh with your blog personality and that might not mesh with the face you put on when you are using foursquare.

    but to Facebook there all part of the same personality – and it’s true, they are, but all of them might not be appropriate for everyone you interact with.

  4. John – I think unfortunately you’re still in the minority. Not enough people really give a crap about their “true social graph” just yet. I wish they would, because I have lots of great startup company ideas to harness this info and let a person control the different “tiers of friendship” and who they met how and why they share stuff with certain people. But it’s still too cumbersome. No one makes lists on Facebook to control their status updates. It’s one of these things that really takes work, especially if you’ve got 5000 friends – but even for the 20-something year old – how psyched are they to filter through their 500 friends and categorize them. Or build up a social graph elsewhere and figure out how to import it appropriately into Facebook. This “common social graph” dream is the next wave, I just hope enough people care about it that people like Zuck and other wild-eyed startupers like me see a big enough market to build these tools.

  5. Facebook is a good place to see this problem, but it extends well beyond; I expect there will be two solutions:

    1) Open Source or free access friend status management: our friends will be sorted into family, close, school, old, alienated, colleagues, associates, attendees, audience, fans and so on. We will have some input into how individuals are managed, but mostly this will be derived from behaviors.

    2) We will have a new understanding of our own persona that is not contingent on place-dependent privacy status; we won’t expect privacy in the traditional sense and will be comfortable exposing facets of our multiple selves to whomever is interested. As you’d expect the current generation is already there.

  6. I cannot imagine a future where people will be OK with one monolithic corporation both owning and using their identity information.

    Eventually, people will own their own identities and will parse them out as they see fit.

    The key is to completely separate the holding of identity information from the usage of it. There could be one of multiple platforms/utilities that hold identity information that users completely own and control. They then decide where and what part of their identities they parse to what users of their identity.

    Its similar to the credit card model today. Credit cards companies hold the bank and number information and then users parse out where it is used.

  7. I’m not sure if I understand your suggestion…

    Are you suggesting some sort of Personality Genome Project that does.. what? Suggest apps, things to do, or things to share? Alter how people view your profile at the time?

  8. It seems to me that the near-ubiquitous Facebook Like buttons already allow you to create your persona on the fly without visiting Facebook, at least to some extent. If you’re looking at a site / product / article, or whatever, you have the ability to tell Facebook that it’s something you like and that it’s okay to share that fact with friends.

    To take the tracking to the level you’re talking about, it would require a plugin or proxy that sees everything you do, but then you’d be forced to sift through all that data (much of it fluff) to determine what you really wanted to share. It’s hard to imagine a mass audience adopting a system that takes so much work, particularly when most people don’t even bother to use friend groups on Facebook.

    On yet another level, I’m not sure that putting this much effort into an online persona is productive. If we were all spending time curating our own activities, deciding what gets shown to our friends, family, colleagues, and strangers, it’s that much less time that we have to actually do things and talk to people. Our acquaintances, friends, and family know us because we talk (either online or in person), not because they read our profile, so is there a purpose in building such a robust identity?

  9. It is true that there is interaction between Facebook and potential customers, but we go on Facebook have fun, read some gossip, chatting with “friends” from elementary school. No one does desire to be spamed at the fun time with advertising and raped by the party invitations from other side of state. What is popular is not necessarily good.

  10. For me, this all comes down to a basic issue of “negotiated identity”. More and more, different providers are going to be able to offer genuine value through insights into (some of) your social graph — filters and serendipitous discoveries that really matter to you.
    On the other hand, that doesn’t mean that you’d ever want to just hand over your entire social graph to any provider you want to check out on a trial basis. Until we work out some kind of straightforward way for to say “You can ‘know’ this much about me to help me out, but nothing else”, this issue is always going to be acting as a brake on any kind of identity economy.

  11. I’m pretty sure old Rishi Dave is on to a central element:

    “The key is to completely separate the holding of identity information from the usage of it. There could be one of multiple platforms/utilities that hold identity information that users completely own and control. They then decide where and what part of their identities they parse to what users of their identity.”

    And I believe the core utility that should hold that information and which users should own and control… is something I’ve been thinking about for a while… and it has a relation to another issue Rishi Dave mentioned… payment systems…

    (Big points to Rishi Dave)

    I believe there are certain transactions that MUST be simplified and unburdened… and in accomplishing that other social benefits arise…

    One of which being the establishment of just the sort of utility you mention:

    The Commons-dedicated Account & Network

    Patent #7,870,067 just granted by USPTO 01/11/’11

    A neutral network of accounts for political, charitable and speech related monetary participation… which in order to properly network and scale individual capability must allow a viable, one-button, secure and financially unburdened micro-transaction. Such a network ideally should maintain its own cloud and bank. Accounts may be created and/or maintained with zero balances and/or only momentary balances during a pass-through transfer (monetization model requires no burden on the actual transaction.)

    From user’s perspective it’s similar to Facebook credits or X-box points except for Commons-oriented functions instead of games, etc. …and most critically NOT adding to or drawing from transaction costs.

    Demo and FAQ
    Civilization Systems Blog

    Regarding its potential in politics, especially for networked citizen lobbying and as an ESSENTIAL tool for citizens participation:

    Addressing Two Problems:

    1. Simplifying the process: “…[campaign contribution] could become much easier if campaigns can figure out how to allow people to donate by making the process easier through one click pay methods and a short form for the additional FEC information required.” – Katie Harbath, chief digital strategist at the National Republican Senatorial Committee… now with Facebook

    2. The need for a viable tool for voter response to the Citizens United Decision and broadening the citizen’s capability for participation in general.


    For simple illustration of the principal:

    Some round figures: $25 times 150 million voters is $3,750,000,000.
    That would be all of 50 cents a week (giving you even a couple of weeks off for holidays!)

    Now I’m not suggesting that it would work exactly that way… but I am suggesting there’s a whole lot more potential there if it can be tapped… if it can be made easy enough that it becomes part of your regular life…

    I believe that CAN and MUST happen.

    The potential for additional monetization of news and journalism is synergistic icing on this cake.

    While a very pragmatic model… its roots rest in a view on the roots of transaction in human society.

    Anybody wanna do something useful?

    P.S. I’ve pretty much abandoned the ICT type blogs but making headway with some of the economics/political types… its just a matter of time and opportunity. Though frankly the planet’s running short of both.

    On the Birth of the Global Social Organism

    Finding Roots in a Shifting Landscape: Facebook and the Future of Social Networks

    Decision Technologies: Currencies and the Social Contract

  12. To be fair… it may be that I haven’t been entirely clear.

    As concisely as possible:

    Dear Mr. Battelle and Mr. O’Reilly,

    You guys are the real deal in the whole ICT/Web 2.0/Gov 2.0 etc sector. You put together the events that make for the necessary face-to-face that finally gets things done.

    I believe this needs to get done:

    Political monetary participation must be unencumbered and simplified (including especially at the micro-transaction level), its networking must be facilitated and the utility that enables that must be (in some form) universally owned. I also believe it can and should be self-supporting and governed by its users.

    I believe the method and model are in rough form noted above but am certainly open to alternative formulations.

    I hope that you will at some point find it worthy of consideration for serious discussion and response.

    Actually I think its inevitable, but the time lag is becoming problematic.

    Interestingly (though not surprisingly) if all else fails this will likely arise either via efforts of frustrated Progressive and/or Libertarian-Tea Party elements in this country. They tend to be the ones I find that get most excited by the concept.

    And then the political establishment will have no choice but to acquiesce out of embarrassment if nothing else.

    I’m just trying to get it built right and obviously can’t do that by my little lonesome.

  13. To take the tracking to the level you’re talking about, it would require a plugin or proxy that sees everything you do, but then you’d be forced to sift through all that data (much of it fluff) to determine what you really wanted to share. It’s hard to imagine a mass audience adopting a system that takes so much work, particularly when most people don’t even bother to use friend groups on Facebook.

  14. @brian – I am in the minority, but I think consumer behavior is going to change and quick, once the value proposition is obvious. See my post on “digital plumage”

  15. You bring up an interesting conundrum – but I don’t think it is answered or caused by the closed system Facebook has or any miracle an open source system can provide.

    In the search realm Google has given us the ability to remove sites we do not want in our search results and no doubt will follow it soon with suggestions of others we may not like – using their similar site option.

    Perhaps we need to create these personas ourselves with a platform like Tweetdeck that lets us swap between accounts. If I want to see information related to my sports interests I can go that way – or maybe they recognize it for me once I make my first site decision. The people I interact with during that ‘site set’ become my ‘social circle’ for that occasion similar to the real world where I have friends that I go to the pub and watch the games with versus those I spend time discussing business or the kids or any of the numerous social aspects of me.

    With behavioral targeting we can get ads sent to us, so the system has to be able to be applied to our social circles. We should be able to tell the system we don’t want to see an ad anymore after it annoyingly follows us, so why not be able to have an area where people outside our current interest circle can attempt to reach us and like caller id be able to pick and choose who and how we engage with them.

    Create that and then the relevance of many things improves.

  16. I don’t know why we all get our knickers in s knot whe it comes down to the web and our so called privacy. The minute we decide to create a profile, whether it be web 2.0, forum, comment, build a website, anything, we have the choice to reveal who we are and how much of what we are. Let’s face it, how many billion people are on the web? Does anyone really have the resources to follow or care or consume themselves with everyone? I am in the business of being interlinked. I have to think each time I trweet if that is ‘real’ me’ or ‘professional’ me, and if so, which account posts where. Let’s face it, big brother today now has names and faces attached to our identities (if we choose). We were always tracked by numbers before, i.e. SIN #, Social security number, credit card numbers, credit rating etc… If people see privacy as a problem, then simple, don’t go online.

  17. Got some good input from someone with experience in areas I’m weakest regarding my concept and model I briefly describe above… (great to get feedback… came via someone over at NakedCapitalism!)

    While I’ve been focusing on need for and benefits of such a system in terms of better governance… and reference to the ‘pre-currency’ fundamentals of transaction…

    In terms of “startup world” (not my specialty), its best to view as an Internet ‘wallet system’ good for ALL types of transactions but that the system can significantly reduce and in some cases eliminate middleman transaction costs because of its monetization potentials in campaign/charity services, advertising and the opportunities offered by the ubiquitous nature of its market and dedicated patented functions.

    Anyway, just a thought… could be just gobbledygook… but wanted to pass it on.

  18. I chose to opt out of the Pandora/Facebook integration for two reasons: 1) I like how the music genome project works (pick a song or artist and they’ll find something similar) and 2) I tune my Pandora stations VERY finely and don’t want the stuff my FB friends “like” to contaminate that tuning.

    When I saw the integration, I gave it a try but when Pandora started playing songs in my stations based on what my friends liked, I was out. That’s rather antithetical to the music genome project purpose. The idea wasn’t to be social music radio, it was to discover music based on music you already like. And it works well. My muddy the waters by going “social”?

Leave a Reply

Your email address will not be published. Required fields are marked *